关中地区地处渭河盆地，近年来社会经济快速发展导致大气问题突出，因此进行空气质量预报预警工作以获取准确及时的大气污染信息也显得尤为重要。相比于传统的统计预报，区域空气质量模式以大气动力学为基础，考虑多种物理和化学过程，进而定量描述污染物的扩散和输送规律，被广泛应用于空气质量预报以及污染过程分析中。但由于现阶段的排放清单具有时间滞后性与低空间分辨率性，且模式本身化学机理以及气象模式也存在不确定性等问题，导致空气质量模式的模拟结果与实际观测值之间存在误差。为了提高关中地区污染物浓度预测精度，本研究针对区域空气质量模型现存问题，结合多模式数值预报与机器学习算法，开展关中地区区域空气质量模式优化与订正方法研究。本文主要研究内容总结如下：

(1)基于大气化学数值模式建立关中地区空气质量多模式预报预警系统。利用WRF、CMAQ及CAMx模式进行关中地区气象数值模拟以及空气质量预报及再分析工作，获取关中地区的数值模拟再分析数据以及WRF模式气象数据，并对WRF、CMAQ及CAMx模式进行相关精度验证。结果表明，WRF模式能准确模拟区域气象因子的时空演变，对温度以及海平面气压的模拟精度较高，相关性系数分别为0.94、0.91；对风速及相对湿度的模拟精度稍低。CAMx与CMAQ模式模拟结果与观测值时序变化趋势一致，模拟结果存在PM_2.5高估、O₃低估现象，其中CAMx模拟精度较CMAQ低。

(2)利用机器学习算法实现单站点模拟结果优化以及优化评价。由于目前大多数研究集中在单个机器学习模型对预报值进行优化，而对多种机器学习模型优化结果的对比以及模型对不同污染物优化效果评估的研究相对较少。因此基于空气质量模式模拟结果，本文提出利用多种机器学习算法对数值模拟结果进行订正与优化，选取温度、相对湿度、边界层高度、海平面气压以及风速等气象因子，验证多种机器学习算法对于关中五市的数值模拟优化结果，分析不同算法对于单站数值模拟的优化性能。结果显示，随机森林算法对PM_2.5优化精度最高，优化后模拟值与观测值之间的相关性系数为0.74~0.8；而支持向量机算法对O₃优化结果最好，优化后相关性系数提高至0.79~0.88。

(3)建立区域优化网络，实现关中地区区域模拟结果优化并进行验证。由于当前机器学习算法研究主要集中在城市站点优化方面，而利用算法结合集合预报实现数值模型的区域优化较少。为实现数值模式的区域优化，本文在单站优化的基础上，加入关中五市33个国控监测站点数据、DEM及土地利用数据，搭建关中地区区域优化网络，验证对比XGBoost算法与LSTM深度学习神经网络算法的优化性能。研究结果显示，XGBoost 算法在区域优化方面优于LSTM算法，XGBoost与LSTM算法优化后的PM_2.5模拟值与观测值相关性系数分别为0.93~0.99、0.73~0.82，O₃模拟值与观测值相关性系数分别0.85~0.97、0.7~0.84。在区域验证方面，算法能明显改善模式PM_2.5高估以及O₃低估的现象，并为区域提供较为准确的污染物浓度值。

(4)利用XGBoost算法实现关中地区未来七天的区域数值模拟优化预报。基于算法区域优化的可行性，本文利用算法实现对区域预报数据的订正优化。结果表明，关中地区PM_2.5预报优化结果的RMSE值约为14.2~22.3μg·m^-3，O₃预报优化结果的RMSE值约为5.5~10.6μg·m^-3，同时PM_2.5 及O₃的浓度演变过程与单站观测结果时序变化结果一致。因此，通过算法优化，可以提高对未来空气质量预报的准确度，对于没有空气质量监测站点的地区也可以实现大气污染物的准确预报，进而为污染管控及防治提供参考。

Guanzhong area is located in the Weihe Basin. In recent years, the rapid development of social economy has led to prominent atmospheric problems. Therefore, it is particularly important to carry out air quality forecast and early warning to obtain accurate and timely air pollution information. Compared with the traditional statistical forecast, the regional air quality model is based on atmospheric dynamics, considering a variety of physical and chemical processes, and then quantitatively describes the diffusion and transport laws of pollutants, which is widely used in air quality forecast and pollution process analysis. However, due to the time lag and low spatial resolution of the emission inventory at the present stage, as well as the uncertainties of the chemical mechanism of the model and the meteorological model, there are errors between the simulated results of the air quality model and the actual observed values. In order to improve the prediction accuracy of pollutant concentration in Guanzhong region, this study aimed at the existing problems of regional air quality model in Guanzhong region, combined with multi-model numerical prediction and machine learning algorithm, carried out optimization and correction method research of regional air quality model. The main research contents of this paper are summarized as follows:

(1) A multi-model air quality forecast and early warning system in Guanzhong region was established based on atmospheric chemistry numerical model. In this paper, WRF, CMAQ and CAM_X models were used to carry out meteorological numerical simulation and air quality forecast and reanalysis in Guanzhong region. The numerical simulation reanalysis data and WRF model meteorological data in Guanzhong region were obtained, and the related accuracy of WRF, CMAQ and CAM_X models was verified. The results show that the WRF model can accurately simulate the spatiotemporal evolution of regional meteorological factors, and the simulation accuracy of temperature and sea level pressure were relatively high, with correlation coefficients of 0.94 and 0.91, respectively. The simulation accuracy of wind speed and relative humidity were lower. The simulation results of CAM_X and CMAQ modes were consistent with the observation values in time series, and PM_2.5 overestimation and O₃ underestimation existed in the simulation results. The simulation accuracy of CAM_X was lower than that of CMAQ.

(2) The machine learning algorithm was used to optimize the simulation results of single site and optimize the evaluation. At present, most of the researches focused on the optimization of forecast values by a single machine learning model, while there were relatively few researches on the comparison of optimization results of multiple machine learning models and the evaluation of optimization effects of different pollutants by models. Therefore, based on the air quality model simulation results, this paper proposed to use a variety of machine learning algorithms to correct and optimize the numerical simulation results. Meteorological factors such as temperature, relative humidity, boundary layer height, sea level pressure and wind speed were selected to verify the numerical simulation optimization results of a variety of machine learning algorithms for the five cities in Guanzhong, and to analyze the optimization performance of different algorithms for the numerical simulation of a single station. The results showed that the random forest algorithm had the highest optimization accuracy for PM_2.5, and the correlation coefficient between the simulated value and the observed value after optimization was 0.74~0.8. The support vector machine algorithm had the best result for O₃ optimization, and the correlation coefficient was improved to 0.79~0.88 after optimization.

(3) The regional optimization network was established to optimize and verify the simulation results in Guanzhong region. At present, the research of machine learning algorithms mainly focused on the optimization of urban sites, while the regional optimization of numerical models based on algorithm combined with ensemble prediction was less. In order to realize the regional optimization of numerical model, based on the single station optimization, this paper added the data of 33 state-controlled monitoring stations, DEM and land use data in five cities of Guanzhong to build a regional optimization network, and verified and compared the optimization performance of XGBoost algorithm and LSTM deep learning neural network algorithm. The results showed that the XGBoost algorithm was superior to the LSTM algorithm in terms of regional optimization. The correlation coefficients between the simulated and observed PM_2.5 values optimized by XGBoost and LSTM algorithms were 0.93-0.99 and 0.73-0.82, respectively. The correlation coefficients between the simulated and observed values of O3 were 0.85~0.97 and 0.70 ~0.84, respectively. In terms of regional verification, the algorithm could significantly improve the phenomenon of overestimation of PM_2.5 and underestimation of O₃, providing more accurate pollutant concentration value for the region.

(4) The XGBoost algorithm was used to achieve the regional numerical simulation and optimization forecast of Guanzhong region in the next seven days. Based on the feasibility of regional optimization algorithm, this paper used the algorithm to revise and optimize the regional forecast data. The results showed that the RMSE value of PM_2.5 prediction and optimization results in Guanzhong area was about 14.2~22.3μg·m^-3, and the RMSE value of O₃ prediction and optimization results was about 5.5~10.6μg·m^-3. Meanwhile, the evolution process of PM_2.5 and O₃ concentration were consistent with the time series variation results of single station observation results. Therefore, the algorithm optimization could improve the accuracy of air quality forecast in the future, and could also achieve accurate forecast of air pollutants in areas without air quality monitoring stations, thus providing reference for pollution control and prevention.

[1] Liu X, Bai X, Tian H, et al. Fine particulate matter pollution in North China: Seasonal-spatial variations, source apportionment, sector and regional transport contributions[J]. Environmental research, 2020, 184: 109368.

[2] Xiao C, Chang M, Guo P, et al. Analysis of air quality characteristics of Beijing–Tianjin–Hebei and its surrounding air pollution transport channel cities in China[J]. Journal of Environmental Sciences, 2020, 87: 213-227.

[3] Wang F, Qiu X, Cao J, et al. Policy-driven changes in the health risk of PM2.5 and O3 exposure in China during 2013–2018[J]. Science of The Total Environment, 2021, 757: 143775.

[4] Lu X, Zhang L, Chen Y, et al. Exploring 2016–2017 Surface Ozone Pollution over China: Source Contributions and Meteorological Influences[J]. Atmospheric Chemistry and Physics, 2019, 19(12): 8339–8361.

[5] Quan J, Xu X, Jia X, et al. Multi-scale processes in severe haze events in China and their interactions with aerosols: Mechanisms and progresses[J]. Chinese Science Bulletin, 2020, 65(9): 810-824.

[6] Xu Y, Xue W, Lei Y, et al. Spatiotemporal variation in the impact of meteorological conditions on PM2.5 pollution in China from 2000 to 2017[J]. Atmospheric Environment, 2020, 223: 117215.

[7] Tai A P K, Mickley L J, Jacob D J. Correlations between fine particulate matter (PM2. 5) and meteorological variables in the United States: Implications for the sensitivity of PM2. 5 to climate change[J]. Atmospheric environment, 2010, 44(32): 3976-3984.

[8] Huang Y L, Yang X. Influence of fine particulate matter on atmospheric visibility[J]. Chinese Science Bulletin, 2013, 58(13): 1165-1170.

[9] Zhu B, Chen R, Xu J X, et al. Health effects of ambient ultrafine (nano) particles in haze[J]. Chinese Science Bulletin, 2015, 60(30): 2808-2823.

[10] 张小曳,孙俊英,王亚强,等. 我国雾-霾成因及其治理的思考[J]. 科学通报, 2013, 58(13): 1178-1187.

[11] 生态环境部. 2019 年中国生态环境状况公报[R]. 北京: 生态环境部, 2020.

[12] 陕西省统计局. 陕西省统计年鉴(2019年)[M]. 北京: 中国统计出版社, 2020.

[13] 陕西省环保厅. 2019年陕西省生态环境状况公报. 2020.

[14] Quan J, Liu Q, Li X, et al. Effect of heterogeneous aqueous reactions on the secondary formation of inorganic aerosols during haze events[J]. Atmospheric Environment, 2015, 122: 306-312.

[15] Guo S, Hu M, Zamora M L, et al. Elucidating severe urban haze formation in China[J]. Proceedings of the National Academy of Sciences, 2014, 111(49): 17373-17378.

[16] Elser M, Huang R J, Wolf R, et al. New insights into PM2.5 chemical composition and sources in two major cities in China during extreme haze events using aerosol mass spectrometry[J]. Atmospheric Chemistry and Physics, 2016, 16(5): 3207-3225.

[17] Wang G, Zhang F, Peng J, et al. Particle acidity and sulfate production during severe haze events in China cannot be reliably inferred by assuming a mixture of inorganic salts[J]. Atmospheric Chemistry and Physics, 2018, 18(14): 10123-10132.

[18] Wang L, Lal V, Khalizov A F, et al. Heterogeneous chemistry of alkylamines with sulfuric acid: Implications for atmospheric formation of alkylaminium sulfates[J]. Environmental science & technology, 2010, 44(7): 2461-2465.

[19] Pai T Y, Lin S H, Yang P Y, et al. Predicting hourly ozone concentration time series in Dali area of Taichung city based on seven types of GM (1, 1) model[M]//Time Series Analysis, Modeling and Applications. Springer, Berlin, Heidelberg, 2013: 369-383.

[20] Koo J W, Wong S W, Selvachandran G, et al. Prediction of Air Pollution Index in Kuala Lumpur using fuzzy time series and statistical models[J]. Air Quality, Atmosphere & Health, 2020, 13(1): 77-88.

[21] Fong I H, Li T, Fong S, et al. Predicting concentration levels of air pollutants by transfer learning and recurrent neural network[J]. Knowledge-Based Systems, 2020, 192: 105622.

[22] Ma J, Cheng J C P, Lin C, et al. Improving air quality prediction accuracy at larger temporal resolutions using deep learning and transfer learning techniques[J]. Atmospheric Environment, 2019, 214: 116885.

[23] Le V D, Bui T C, Cha S K. Spatiotemporal deep learning model for citywide air pollution interpolation and prediction[J]. arXiv preprint arXiv:1911.12919, 2019.

[24] Pak U, Ma J, Ryu U, et al. Deep learning-based PM2.5 prediction considering the spatiotemporal correlations: A case study of Beijing, China[J]. Science of The Total Environment, 2020, 699: 133561.

[25] 万显烈,杨凤林,王慧卿.利用人工神经网络对空气中O3浓度进行预测[J].中国环境科学,2003(01):111-113.

[26] 周杉杉,李文静,乔俊飞.基于MRMR-PSO的递归模糊神经网络PM2.5预测模型[J].计算机与应用化学,2018,35(10):783-791.

[27] 康俊锋,黄烈星,张春艳,等.多机器学习模型下逐小时PM2.5预测及对比分析[J].中国环境科学,2020,40(05):1895-1905.

[28] 曲悦,钱旭,宋洪庆,等.基于机器学习的北京市PM2.5浓度预测模型及模拟分析[J].工程科学学报,2019,41(03):401-407.

[29] Li T, Shen H, Yuan Q, et al. Estimating ground‐level PM2. 5 by fusing satellite and station observations: a geo‐intelligent deep learning approach[J]. Geophysical Research Letters, 2017, 44(23): 11,985-11,993.

[30] Li T, Zhang C, Shen H, et al. Real-time and Seamless Monitoring of Ground-Level PM2.5 Using Satellite Remote Sensing[J]. arXiv preprint arXiv:1803.03409, 2018.

[31] Kianian B, Liu Y, Chang H H. Imputing Satellite-Derived Aerosol Optical Depth Using a Multi-Resolution Spatial Model and Random Forest for PM2.5 Prediction[J]. Remote Sensing, 2021, 13(1): 126.

[32] Lee H J, Coull B A, Bell M L, et al. Use of satellite-based aerosol optical depth and spatial clustering to predict ambient PM2. 5 concentrations[J]. Environmental research, 2012, 118: 8-15.

[33] Wang X, Sun W, Zheng K, et al. Estimating hourly PM2.5 concentrations using MODIS 3 km AOD and an improved spatiotemporal model over Beijing-Tianjin-Hebei, China[J]. Atmospheric Environment, 2020, 222: 117089.

[34] Jie C, Zhengqiang L I, Wenyuan C, et al. Impact Evaluation of Aerosol Variational Assimilation Based on Improved GSI System on WRF-Chem PM2.5 Analysis and Forecast[J]. Journal of Atmospheric and Environmental Optics, 2020, 15(05): 321.

[35] Ha S, Liu Z, Sun W, et al. Improving air quality forecasting with the assimilation of GOCI aerosol optical depth (AOD) retrievals during the KORUS-AQ period[J]. Atmospheric Chemistry and Physics, 2020, 20(10): 6015-6036.

[36] Zhang H, Cheng S, Yao S, et al. Multiple Perspectives for Modeling Regional PM2.5 Transport across Cities in the Beijing–Tianjin–Hebei Region during Haze Episodes[J]. Atmospheric Environment, 2019, 212: 22–35.

[37] Jiang H, Geng X, Cheng Z, et al. Source apportionment of PM2.5 carbonaceous aerosols during a long-lasting winter haze episode in Xiangyang, central China[J]. Atmospheric Pollution Research, 2021, 12(2): 470-479.

[38] Lin J, An J, Qu Y, et al. Local and Distant Source Contributions to Secondary Organic Aerosol in the Beijing Urban Area in Summer[J]. Atmospheric Environment, 2016, 124: 176–185.

[39] Li R, Mei X, Wei L, et al. Study on the Contribution of Transport to PM2.5 in Typical Regions of China Using the Regional Air Quality Model RAMS-CMAQ[J]. Atmospheric Environment, 2019, 214: 116856.

[40] Chen D, Liu X, Lang J, et al. Estimating the Contribution of Regional Transport to PM 2.5 Air Pollution in a Rural Area on the North China Plain[J]. Science of The Total Environment, 2017, 583: 280–291.

[41] 薛文博,王金南,杨金田,等.国内外空气质量模型研究进展[J].环境与可持续发展,2013,38(03):14-20.

[42] 王卓. 国内外空气质量模型研究与应用进展[A]. 中国环境科学学会(Chinese Society For Environmental Sciences).2015年中国环境科学学会学术年会论文集（第一卷,2015:6.

[43] N.S. Holmes, L. Morawska. A review of dispersion modelling and its application to the dispersion of particles: An overview of different dispersion models available[J]. Atmospheric Environment,2006,40(30).

[44] 胡翠娟,丁峰,李时蓓,等.国内外环境空气质量模型法规化现状与对比研究[J].环境工程,2015,33(01):132-136.

[45] 王淑芳,陈家宜,刘式达.北京城近郊区大气污染评价的气质模式[J]. 中国环境科学, 1981, 1(1):0-0.

[46] 王自发,谢付莹,王喜全,等.嵌套网格空气质量预报模式系统的发展与应用[J].大气科学,2006(05):778-790.

[47] 王自发,吴其重,晏平仲,等.北京空气质量多模式集成预报系统的建立及 初步应用[J]. 南京信息工程大学学报: 自然科学版, 2009, 1(1): 19-26.

[48] 黄思,唐晓,徐文帅,等.利用多模式集合和多元 线性回归改进北京PM10预报[J].环境科学学报,2015,35(01):56-64.

[49] 程兴宏,刁志刚,胡江凯,等.基于CMAQ模式和自适应偏最小二乘回归法的中国地区PM2.5浓度动力-统计预报方法研究[J].环境科学学报,2016,36(8):2771-2782.

[50] 赵俊日,肖昕,吴涛,等.空气质量数值预报优化方法研究[J].中国环境科学,2018,38(6):2047-2054.

[51] 汤静,王春林,谭浩波,等.利用PCA-kNN方法改进广州市空气质量模式PM2.5预报[J].热带气象学报,2019,35(1):125-134.

[52] 张伟,王自发,安俊岭,等.利用BP神经网络提高奥运会空气质量实时预报系统预报效果[J].气候与环境研究,2010,15(5):595-601.

[53] 王茜,吴剑斌,林燕芬.CMAQ模式及其修正技术在上海市PM2.5预报中的应用检验[J].环境科学学报,2015,35(6):1651-1656.

[54] 张岳军,张怀德,朱凌云,等.太原市PM2.5预报统计修正模型及其应用检验[J].环境科学研究,2018,31(7):1207-1213.

[55] 潘锦秀,晏平仲,孙峰,等.多元线性回归方法对北京地区PM2.5预报的改进应用[J].中国环境监测,2019,35(2):43-52.

[56] 张恒德,张庭玉,李涛,等.基于BP神经网络的污染物浓度多模式集成预报[J].中国环境科学,2018,38(4):1243-1256.

[57] 戴李杰,张长江,马雷鸣.基于机器学习的PM2.5短期浓度动态预报模型[J].计算机应用,2017,37(11):3057-3063.

[58] 马井会,曹钰,余钟奇,等.深度学习方法在上海市PM2.5 浓度预报中的应用[J].中国环境科学,2020,40(2):530-538.

[59] 陕西省地方志编纂委员会. 陕西省志·地理志[M]. 西安: 陕西人民出版社, 2000.

[60] 生态环境部.环境空气质量指数(AQI)技术规定(试行)(633-2012)[S]. 北京: 生态环境部, 2012.

[61] Guenther A, Karl T, Harley P, et al. Estimates of global terrestrial isoprene emissions using MEGAN (Model of Emissions of Gases and Aerosols from Nature)[J]. Atmospheric Chemistry and Physics, 2006, 6(11): 3181-3210.

[62] Maindonald J, Braun J. Data analysis and graphics using R: an example-based approach[M]. Cambridge University Press, 2006.

[63] 李颖若,汪君霞,韩婷婷,等.利用多元线性回归方法评估气象条件和控制措施对APEC期间北京空气质量的影响[J].环境科学,2019,40(03):1024-1034.

[64] Breiman, L. Random Forests, Mach. Learn., 45, 5–32.

[65] Keller C A, Evans M J. Application of random forest regression to the calculation of gas-phase chemistry within the GEOS-Chem chemistry model v10[J]. Geoscientific Model Development, 2019, 12(3): 1209-1225.

[66] Suykens J A K, Vandewalle J. Least squares support vector machine classifiers [J]. Neural Processing Letters, 1999,9(3):293-300.

[67] 张学工.关于统计学习理论与支持向量机 [J]. 自动化学报, 2000, 26(1):36-46.

[68] Chen,T., Tong,H., Benesty,M. Xgboost: Extreme Gradient Boosting[C]. Proceedings of ACM Sig KDD International onference on Knowledge Discovery and Data Mining.2016,85-794.

[69] Gers F A, Schmidhuber J, Cummins F. Learning to forget: Continual prediction with LSTM[J]. 1999.

[70] Li T, Hua M, Wu X. A hybrid CNN-LSTM model for forecasting particulate matter (PM2. 5)[J]. IEEE Access, 2020, 8: 26933–26940.

[71] Boylan J W, Russell A G. PM and light extinction model performance metrics, goals,and criteria for three-dimensional air quality models[J].Atmospheric Environment,2005,40

:4946-4959.

[72] Bai L, Wang J, Ma X, et al. Air Pollution Forecasts: An Overview[J]. International Journal of Environmental Research and Public Health, 2018, 15(4): 780.

[73] Pontoppidan M, Reuder J, Mayer S, et al. Downscaling an intense precipitation event in complex terrain: The importance of high grid resolution[J]. Tellus A: Dynamic Meteorology and Oceanography, 2017, 69(1): 1271561.

[74] Carvalho D, Rocha A, Gómez-Gesteira M, et al. A sensitivity study of the WRF model in wind simulation for an area of high wind energy[J]. Environmental Modelling & Software, 2012, 33: 23–34.

[75] Powers J G, Klemp J B, Skamarock W C, et al. The weather research and forecasting model: Overview, system efforts, and future directions[J]. Bulletin of the American Meteorological Society, 2017, 98(8): 1717–1737.

[76] Gómez-Navarro J J, Raible C C, Dierer S. Sensitivity of the WRF Model to PBL Parametrisations and Nesting Techniques: Evaluation of Wind Storms over Complex Terrain[J]. Geoscientific Model Development, 2015, 8(10): 3349–3363.

[77] 颜鹏,李维亮,秦瑜.近年来大气气溶胶模式研究综述[J].应用气象学报,2004(05):629-640.

[78] Zheng B, Zhang Q, Zhang Y, et al. Heterogeneous chemistry: a mechanism missing in current models to explain secondary inorganic aerosol formation during the January 2013 haze episode in North China[J]. Atmospheric Chemistry and Physics, 2015, 15(4): 2031-2049.

[79] 陈云波,徐峻,何友江,等.北京市冬季典型重污染时段PM2.5污染来源模式解析[J].环境科学研究,2016,29(05):627-636.

[80] 何斌,梅士龙,陆琛莉,等.MEIC排放清单在空气质量模式中的应用研究[J].中国环境科学,2017,37(10):3658-3668.

[81] Li J, Han Z, Wu Y, et al. Aerosol radiative effects and feedbacks on boundary layer meteorology and PM2.5 chemical components during winter haze events over the Beijing-Tianjin-Hebei region[J]. Atmospheric Chemistry and Physics, 2020, 20(14): 8659-8690.

[82] 沈劲,王雪松,李金凤,等.Models-3/CMAQ和CAMx对珠江三角洲臭氧污染模拟的比较分析[J].中国科学:化学,2011,41(11):1750-1762.

[83] Wang X, Wei W, Cheng S, et al. Characteristics and classification of PM2.5 pollution episodes in Beijing from 2013 to 2015[J]. Science of the Total Environment, 2018, 612: 170-179.

[84] Honomichl S B, Pan L L. Transport from the Asian summer monsoon anticyclone over the western Pacific[J]. Journal of Geophysical Research: Atmospheres, 2020, 125(13): e2019JD032094.

[85] Yuan, C, He K B, Du Z Y, et al. Humidity plays an important role in the PM2.5 pollution in Beijing[J]. Environmental Pollution, 2015,197:68-75.

[86] Li M, Wang L, Liu J, et al. Exploring the Regional Pollution Characteristics and Meteorological Formation Mechanism of PM2.5 in North China during 2013–2017[J]. Environment International, 2020, 134: 105283.

[87] Jia W, Zhang X, Wang Y. Assessing the Pollutant Evolution Mechanisms of Heavy Pollution Episodes in the Yangtze-Huaihe Valley: A Multiscale Perspective[J]. Atmospheric Environment, 2021, 244: 117986.

[88] Huang R J, Duan J, Li Y, et al. Effects of NH3 and alkaline metals on the formation of particulate sulfate and nitrate in wintertime Beijing[J]. Science of The Total Environment, 2020, 717: 137190.

[89] Liu P, Ye C, Xue C, et al. Formation Mechanisms of Atmospheric Nitrate and Sulfate during the Winter Haze Pollution Periods in Beijing: Gas-Phase, Heterogeneous and Aqueous-Phase Chemistry[J]. Atmospheric Chemistry and Physics, 2020, 20(7): 4153–4165.

[90] Tian M, Liu Y, Yang F, et al. Increasing importance of nitrate formation for heavy aerosol pollution in two megacities in Sichuan Basin, southwest China[J]. Environmental Pollution, 2019, 250: 898–905.

[91] Mao J, Wang L, Lu C, et al. Meteorological mechanism for a large-scale persistent severe ozone pollution event over eastern China in 2017[J]. Journal of Environmental Sciences, 2020, 92: 187–199.

[92] Otero N, Rust H W, Butler T. Temperature dependence of tropospheric ozone under NOx reductions over Germany[J]. Atmospheric Environment, 2021: 118334.

[93] Liu P, Song H, Wang T, et al. Effects of meteorological conditions and anthropogenic precursors on ground-level ozone concentrations in Chinese cities[J]. Environmental Pollution, 2020, 262: 114366.

[94] 李娟,尉鹏,戴学之,等.基于机器学习方法的西安市数值模拟优化研究[J].环境科学研究.2021,34(4):34-43.

[95] ENVIRON. User’s Guide Comprehensive Air Quality Model With Extensions[M]. Version 7.0.Novato,California:Ramboll Environment and Health.2020.

[96] Li R, Wang Z, Cui L, et al. Air pollution characteristics in China during 2015–2016: Spatiotemporal variations and key meteorological factors[J]. Science of The Total Environment, 2019, 648: 902–915.

[97] Jin J Q, Du Y, Xu L J, et al. Using Bayesian spatio-temporal model to determine the socio-economic and meteorological factors influencing ambient PM2.5 levels in 109 Chinese cities[J]. Environmental Pollution, 2019, 254: 113023.

[98] Wei P, Cheng S, Li J, et al. Impact of Boundary-Layer Anticyclonic Weather System on Regional Air Quality[J]. Atmospheric Environment, 2011, 45(14): 2453–2463.

[99] Lu D, Mao W, Yang D, et al. Effects of Land Use and Landscape Pattern on PM2.5 in Yangtze River Delta, China[J]. Atmospheric Pollution Research, 2018, 9(4): 705–713.

[100] McCarty J, Kaza N. Urban form and air quality in the United States[J]. Landscape and Urban Planning, 2015, 139: 168–179.

[101] Tang X, Liu H, Li J, et al. Response analysis of haze/particulate matter pollution to land use/cover in Beijing[J]. China Environ. Sci, 2015, 35(9): 2561–2569.

[102] Wu J, Xie W, Li W, et al. Effects of urban landscape pattern on PM2.5 pollution—a Beijing case study[J]. PloS one, 2015, 10(11): e0142449.

[103] Li Y, Liu H, Lou C. Spatial and temporal variation of PM2.5 and influences of land use pattern on PM2.5 concentration in Jiangsu Province[J]. Environ. Sci. Technol, 2016, 39(8): 10–15.

[104] 杜金龙,朱记伟,解建仓,等.近25a关中地区土地利用及其景观格局变化[J].干旱区研究,2018,35(01):217-226.

[105] Lai H C, Lin M C. Characteristics of the upstream flow patterns during PM2.5 pollution events over a complex island topography[J]. Atmospheric Environment, 2020, 227: 117418.

[106] Zhang X, Liu L Y, Wang W J, et al. A SPECLib-based operational classification approach: A preliminary test on China land cover mapping at 30 m[J]. International Journal of Applied Earth Observations and Geoinformation,2018,71.